Two different games, two different rule books
For the past two decades, SEO has meant one thing: rank higher on Google. Get on page one, ideally in the top three results, and the traffic follows. It's a well-understood discipline with established tools, clear metrics, and a predictable feedback loop.
AI visibility is something else entirely. It's not about ranking positions. It's about whether AI systems like ChatGPT, Perplexity, and Google's Gemini include your business in their answers at all. No position one. No position ten. Just mentioned, or not mentioned.
That distinction sounds simple, but the implications run deep. The strategies, signals, and technical foundations that make you visible to AI are substantially different from those that move the needle in traditional search. Some overlap exists, but assuming they're the same will leave you invisible in both places.
What traditional SEO actually optimises for
Classic SEO is built around crawlability, relevance, and authority. Google's crawlers index your pages, assess their content quality, evaluate how other sites link to them, and assign a ranking. You optimise title tags, meta descriptions, headings, and body copy so that Google's algorithm understands what each page is about. You build backlinks to signal authority.
The output is a ranked list. Your page either appears for a given query or it doesn't. If it does, the user still has to click. You're competing for attention on a results page filled with other links, ads, featured snippets, and image carousels.
Performance is measurable in fairly straightforward terms: impressions, clicks, click-through rate, and position. Tools like Google Search Console, Ahrefs, and Semrush make it possible to track what's working and what isn't, often within days of making changes.
Traditional SEO rewards consistency. Update your page, build a few links, wait a few weeks, check your rankings. The system is iterative and, for many businesses, it works reasonably well. The problem is that the system itself is changing.
How AI search engines work differently
AI search engines don't return a ranked list of links. They generate a response. When someone types "what's the best protein powder for women over 40" into Perplexity or asks ChatGPT for a recommendation, the AI constructs an answer using information it has learned, retrieved, or synthesised from multiple sources.
Your website might be one of those sources. Or it might not. The decision isn't based on where you rank in Google. It's based on whether the AI has enough structured, trustworthy, specific information about your business to include you confidently in its answer.
This is a fundamentally different type of visibility. A few things drive it:
- Structured data and schema markup. AI systems process structured information far more efficiently than unstructured prose. If your product pages, reviews, business details, and FAQs are marked up with correct JSON-LD schema, the AI can understand what you sell, who you sell it to, and why you're credible.
- Clear, factual, entity-based content. AI models are trained on and retrieve content that makes clear, attributable claims. Vague marketing copy ("we deliver exceptional experiences") is nearly useless. Specific facts ("we ship within 24 hours to UK mainland addresses, with same-day dispatch on orders placed before 1pm") are exactly what AI systems can pick up and repeat.
- Citation-worthiness. Retrieval-augmented generation (RAG), the technology behind tools like Perplexity and ChatGPT's Browse mode, pulls live web content into responses. Pages that are clearly authoritative on a specific topic, well-structured, and easy to extract information from are far more likely to be cited.
- llms.txt and robots.txt configuration. These files tell AI crawlers what they're allowed to access. Getting this wrong means AI systems can't read your content at all, regardless of how good it is.
You can read more about one of those signals in our guide on what llms.txt is and whether you need it.
The signals that matter in each channel
It helps to put the two approaches side by side.
Traditional SEO signals
- Backlink profile and domain authority
- On-page keyword usage and semantic relevance
- Page speed and Core Web Vitals
- Internal linking structure
- Title tags, meta descriptions, and heading hierarchy
- User engagement metrics (dwell time, bounce rate)
AI visibility signals
- Schema markup completeness and accuracy (Product, Review, FAQ, Organization, BreadcrumbList, and more)
- Entity clarity: does the AI know exactly who you are, what you sell, and where you operate?
- Content specificity: concrete claims, numbers, named products, real policies
- Crawl accessibility for AI bots (Googlebot is not the same as GPTBot or PerplexityBot)
- Structured FAQs that mirror natural language questions
- Brand mentions across trusted third-party sources
The overlap between these two lists is real but limited. A fast, well-linked site with strong on-page SEO is a good foundation. But it won't automatically translate into AI visibility if your structured data is thin, your content is vague, or your site is accidentally blocking AI crawlers.
Why rankings alone no longer measure success
Here's a scenario that's becoming more common. An e-commerce brand ranks on page one of Google for its main product category. Traffic is steady. Conversions are reasonable. Then, quietly, a portion of their potential customers starts using AI tools instead of Google to research purchases. Those customers ask ChatGPT which brand to buy. ChatGPT doesn't mention the brand at all, despite the strong Google ranking, because the brand's product pages have no structured data, no clear reviews schema, and no FAQ content that the AI can extract and cite.
The Google rankings haven't dropped. The SEO metrics look fine. But a growing slice of demand is invisible to them.
This is why measuring AI visibility as a separate metric matters. At FlinnSchema, we track which queries mention a client's brand in AI responses, how often, and in what context. That data doesn't show up in Google Search Console. It requires a different measurement approach entirely. Our post on how to measure AI visibility goes into the practical detail of how that works.
The technical work that underpins AI visibility
One of the most common misconceptions is that AI visibility is purely a content problem. Write better copy, be more specific, and the AI will find you. That's part of it, but the technical layer is just as important.
Schema markup and JSON-LD
Schema markup is the structured data standard that tells machines, including AI systems, exactly what your content means. A page that says "our shoes are great quality" means almost nothing to a machine. A page with correctly implemented Product schema, including price, availability, brand, aggregate rating, and product category, gives AI systems enough structured information to reference your product accurately.
Most e-commerce sites have weak schema. Default Shopify themes include some basic markup, but it's rarely complete or correctly implemented. Missing fields, incorrect types, and outdated syntax are common. These gaps translate directly into reduced AI visibility.
Crawl configuration for AI bots
AI companies run their own crawlers. OpenAI uses GPTBot. Anthropic uses ClaudeBot. Perplexity uses PerplexityBot. These are separate from Googlebot and follow their own rules. If your robots.txt file is blocking these crawlers, intentionally or accidentally, those AI systems simply can't read your site.
This is a surprisingly common issue. Many robots.txt configurations set up to block scrapers also block legitimate AI crawlers. Our detailed guide on robots.txt and AI visibility walks through how to check and fix this.
Content architecture and entity clarity
AI systems build mental models of entities: businesses, products, people, places. The clearer and more consistent your information is across your site, the more confidently an AI can represent you in a response. Your business name, address, product names, brand values, and key claims should be consistent and machine-readable throughout your content and schema.
Do you need to choose between SEO and AI visibility?
No. The two are complementary, not competing. Good technical SEO, strong content, and a healthy backlink profile all contribute to a credible web presence that AI systems can draw on. You're not abandoning one for the other.
But they do require different actions. Traditional SEO work like keyword research, link building, and Core Web Vitals optimisation won't automatically improve your AI visibility. You need additional, specific work: schema implementation, content structured around factual claims, AI crawler configuration, and monitoring of AI search responses.
Thinking of AI visibility as an extension or upgrade to your existing SEO programme is a reasonable framing. It builds on what you've already invested in, but it requires deliberate effort in new directions.
If you're not sure where you currently stand, a free AI visibility audit is a good starting point. It shows you exactly which signals are missing and where the quick wins are.
The shift in how customers search
The practical reason all of this matters is customer behaviour. A growing number of people, particularly in the 25 to 45 age bracket, now start product research with an AI tool rather than a search engine. They ask conversational questions and expect a synthesised answer, not a list of ten blue links to click through.
That behaviour isn't going to reverse. The question for e-commerce brands is whether they're present in those AI-generated answers or absent from them. Getting present requires understanding what AI visibility actually is, how it differs from SEO, and what specific steps close the gap.
Traditional SEO got you in front of customers who were already searching. AI visibility is about being the answer when they ask.
Frequently Asked Questions
Can I improve my AI visibility without touching my SEO?
Yes, to a degree. Schema markup, content specificity, and AI crawler configuration can all be improved without changing your existing SEO setup. In practice, the two programmes work better together, but they can be pursued independently. Fixing your structured data, for example, is purely a technical task that doesn't require changing your keyword strategy or link building.
Will ranking well on Google automatically make me visible in AI search?
Not automatically, no. A high Google ranking means your page is indexed and considered authoritative by Google's algorithm. But AI systems don't simply read Google's rankings. They use their own crawlers, their own training data, and their own criteria for what counts as citable information. Structured data, content specificity, and entity clarity matter independently of where you rank.
How do I know if AI search engines can even access my website?
Check your robots.txt file, typically found at yourdomain.com/robots.txt. Look for any Disallow rules that apply to GPTBot, ClaudeBot, PerplexityBot, or broad wildcards that might inadvertently block them. Our robots.txt guide explains exactly what to look for and how to adjust it safely.
Is AI visibility only relevant for large e-commerce brands?
No. Smaller brands often have an advantage here because their product range is more focused and their content can be more specific. A niche e-commerce brand with well-implemented schema and clear, factual product content can appear in AI responses ahead of large competitors with bloated, vague content. AI systems reward clarity and structure, not just size or authority.